Writers of Pro Football Prospectus 2008

Most Recent FO Features

ProctorSpe15.jpg

» VN: Hits, Misses, and the Year Zero Effect

Bill Connelly looks at the college offenses, defenses, and overall teams that have improved (Air Force!) or regressed (North Texas!) the most in 2014. Year Zero is a real thing (sometimes).

11 Oct 2012

FEI Week 6: Big 12 Taking Charge

by Brian Fremeau

The No. 1 team in the country according to FEI is scoring 52 points per game (fifth most nationally) and 4.2 points per possession (second most nationally). Its offense leads the nation in explosive drive percentage (31.1 percent) and points per ordinary drive (2.7), ranks second in value drive percentage (67.4 percent), and third in available yards percentage (66.4 percent). There’s no dispute that West Virginia has one of the most dangerous offenses in college football.

The Mountaineers have also given up 108 points in the last two games and rank 112th overall in raw defensive efficiency and 111th overall in special teams efficiency. Can West Virginia truly be the nation’s best team if it only plays exceptionally well in only one of three phases of the game? It certainly doesn’t seem likely to be sustainable.

The primary reason West Virginia ranks first overall is tied up in a bit of circular logic. There are five teams from the Big 12 ranked among the top 10 in this week’s FEI ratings (including West Virginia’s last two opponents, No. 6 Texas and No. 9 Baylor) and four others rank among the top 30. Aside from Kansas, conference collectively beat down on its non-conference opponents in such efficient fashion that they are all receiving a collective boost by having played a couple of games apiece against one another. With even more to come –- the conference plays a round-robin schedule -– FEI may continue to reward the league, though I expect there will eventually be some separation.

What’s interesting to me is that while FEI thinks very highly of the league, it doesn’t think very highly of the league’s chance to send a team to the BCS championship game. Only West Virginia and Oklahoma are projected to win more than five more games the rest of the way, meaning FEI thinks that only the Mountaineers will get to double-digit victories by season’s end. FEI says that teams in other conferences have a better chance to get to 10 or more wins: Alabama, Oregon, Florida, Florida State, and Notre Dame.

A very intriguing strength of schedule debate may present itself over the course of the coming months. It seems blasphemous to suggest that an SEC team may not have a schedule strength argument, but Alabama has the 78th-toughest regular season schedule according to current FEI ratings, weaker than traditional AQ conference schedules and much weaker than traditional SEC schedules. Alabama’s dominance of its schedule, however weak, will make this a non-issue, but in a scenario in which 1-loss teams will need to be parsed, the SEC may actually hold the Crimson Tide back.

This is the final week in which preseason data is still used as a component in the FEI ratings. Starting next week, preseason data will be eliminated, and we will debut the complete offensive, defensive, and special teams opponent-adjusted ratings for all teams. Let the arguments begin.

Week 6 Revisionist Box Scores

This weekly feature identifies the games played each week that were most impacted by turnovers, special teams, field position, or some combination of the three. The neutralized margin of victory is a function of the point values earned and surrendered based on field position and expected scoring rates.

Week 6 Games In Which Total Turnover Value Exceeded Non-Garbage Final Score Margin
Date Winning Team Non-Garbage
Final Score
Losing Team TTV
+
TTV
-
TTV
Net
TO Neutral
Score Margin
10/5 BYU 6-3 Utah State 6.8 3.0 3.8 -0.8
10/5 Syracue 14-13 Pittsburgh 7.8 4.8 3.0 -2.0
10/6 Idaho 26-18 New Mexico State 14.3 5.8 8.5 -0.5
10/6 Iowa State 37-23 TCU 23.4 2.1 21.3 -7.3
10/6 Navy 28-21 Air Force 11.9 0.0 11.9 -4.9
10/6 Ohio 38-31 Buffalo 11.8 2.0 9.8 -2.8
10/6 Temple 37-28 South Florida 11.0 0.0 11.0 -2.0
10/6 Tulsa 45-38 Marshall 12.6 3.7 8.9 -1.9

Week 6 Games In Which Special Teams Value Exceeded Non-Garbage Final Score Margin
Date Winning Team Non-Garbage
Final Score
Losing Team STV
+
STV Neutral
Score Margin
10/6 Idaho 26-18 New Mexico State 9.0 -1.0
10/6 Memphis 14-10 Rice 5.3 -1.3
10/6 Ohio 38-31 Buffalo 15.1 -8.1

Week 6 Games In Which Field Position Value Exceeded Non-Garbage Final Score Margin
Date Winning Team Non-Garbage
Final Score
Losing Team FPV
+
FPV
-
FPV
Net
FPV Neutral
Score Margin
10/5 Syracuse 14-13 Pittsburgh 19.4 15.7 3.7 -2.7
10/6 Army 34-31 Boston College 22.0 17.4 4.6 -1.6
10/6 Ohio 38-31 Buffalo 38.6 16.4 22.2 -15.2
10/6 Tulsa 45-38 Marshall 33.9 22.7 11.2 -4.2

2012 totals to date:

  • Net Total Turnover Value was the difference in 47 of 287 FBS games (16.4 percent)
  • Net Special Teams Value was the difference in 22 of 287 FBS games (7.7 percent)
  • Net Field Position Value was the difference in 32 of 287 FBS games (11.1 percent)
  • Turnovers, Special Teams and/or Field Position was the difference in 65 of 287 FBS games (22.6 percent)

2012 Game Splits for all teams, including the offensive, defensive, special teams, field position, and turnover values recorded in each FBS game are provided here.

FEI Week 6 Top 25

The Fremeau Efficiency Index (FEI) rewards playing well against good teams, win or lose, and punishes losing to poor teams more harshly than it rewards defeating poor teams. FEI is drive-based and it is specifically engineered to measure the college game.

FEI is the opponent-adjusted value of Game Efficiency (GE), a measurement of the success rate of a team scoring and preventing opponent scoring throughout the non-garbage-time possessions of a game. FEI represents a team's efficiency value over average. Strength of Schedule (SOS) is calculated as the likelihood that an "elite team" (two standard deviations above average) would win every game on the given team's schedule. SOS listed here includes future games scheduled.

Mean Wins (FBS MW) represent the average total games a team with the given FEI rating should expect to win against its complete schedule of FBS opponents. Remaining Mean Wins (FBS RMW) represent the average expected team wins for games scheduled but not yet played.

Offensive Efficiency (OE) is the raw unadjusted efficiency of the given team's offense, a measure of its actual drive success against expected drive success based on field position. Defensive Efficiency (DE) is the raw unadjusted efficiency of the given team's defense, a measure of the actual drive success of its opponents against expected drive success based on field position. Field Position Advantage (FPA) is the share of the value of total starting field position earned by each team against its opponents.

Only games between FBS teams are considered in the FEI calculations. Since limited data is available in the early part of the season, preseason projections are factored into the current ratings. The weight given to projected data will be reduced each week until Week 7, when it will be eliminated entirely. Opponent-adjusted offensive and defensive FEI ratings will also debut in Week 7.

These FEI ratings are a function of results of games played through October 6. The ratings for all FBS teams can be found here.

Rank Team FBS
W-L
FEI Last
Wk
GE GE
Rk
SOS SOS
Rk
FBS
MW
FBS
RMW
OE OE
Rk
DE DE
Rk
FPA FPA
Rk
1 West Virginia 4-0 .308 10 .164 21 .032 10 8.7 5.4 1.392 1 .495 112 .482 76
2 Alabama 5-0 .304 1 .432 1 .385 78 10.3 5.5 .335 25 -1.009 1 .612 2
3 Notre Dame 5-0 .280 3 .241 6 .111 14 10.3 5.8 .245 34 -.756 5 .515 48
4 Kansas State 4-0 .278 7 .333 3 .016 5 7.9 4.7 .802 4 -.160 44 .593 3
5 Oklahoma 2-1 .260 9 .176 17 .017 6 8.5 6.1 .349 24 -.430 20 .506 55
6 Texas 4-1 .258 2 .229 7 .022 7 8.6 4.6 .887 3 .329 96 .591 5
7 Florida 5-0 .235 5 .182 16 .189 28 9.0 4.8 .031 62 -.524 14 .543 23
8 Florida State 3-1 .235 4 .200 13 .367 72 8.5 5.1 .260 32 -.418 22 .576 7
9 Baylor 2-1 .219 25 .163 22 .009 2 6.4 4.6 1.269 2 1.065 123 .575 8
10 South Carolina 6-0 .213 12 .320 4 .304 56 9.2 3.5 .192 41 -.942 3 .542 24
11 Oregon 5-0 .204 8 .367 2 .259 43 8.9 4.2 .514 12 -.624 6 .535 34
12 Texas Tech 3-1 .198 6 .197 14 .022 8 6.5 3.7 .636 7 -.175 41 .499 62
Rank Team FBS
W-L
FEI Last
Wk
GE GE
Rk
SOS SOS
Rk
FBS
MW
FBS
RMW
OE OE
Rk
DE DE
Rk
FPA FPA
Rk
13 Ohio State 6-0 .190 24 .174 19 .360 69 9.9 4.8 .397 19 -.242 36 .513 49
14 Cincinnati 3-0 .171 11 .264 5 .448 84 7.9 5.2 .386 20 -.580 9 .568 9
15 Texas A&M 3-1 .166 38 .219 9 .164 23 7.0 4.1 .561 10 -.349 27 .542 26
16 Iowa State 3-1 .160 58 .054 54 .023 9 5.4 2.9 -.311 94 -.467 19 .529 38
17 USC 4-1 .154 18 .206 11 .213 31 8.5 4.5 .148 48 -.476 17 .540 27
18 Arizona State 3-1 .152 26 .168 20 .268 47 8.0 4.7 .183 42 -.535 13 .554 19
19 Rutgers 4-0 .149 16 .136 29 .409 81 8.5 5.3 .043 59 -.519 16 .548 21
20 Michigan State 4-2 .145 17 .085 45 .244 40 8.5 4.5 -.108 73 -.520 15 .492 71
21 Oregon State 4-0 .138 32 .072 48 .272 50 7.4 4.5 -.068 68 -.400 26 .498 64
22 Oklahoma State 1-2 .136 27 .070 49 .014 3 5.0 3.1 .747 5 .379 102 .444 111
23 Louisiana Monroe 3-2 .122 41 .133 30 .516 92 10.0 6.6 .677 6 -.009 60 .499 61
24 Michigan 3-2 .121 23 .089 42 .082 12 7.6 4.8 .201 40 -.255 33 .480 80
25 BYU 3-2 .118 59 .144 26 .259 42 8.0 3.9 -.227 89 -.992 2 .500 59

Posted by: Brian Fremeau on 11 Oct 2012

8 comments, Last at 12 Oct 2012, 5:50pm by cfn_ms

Comments

1
by Kal :: Thu, 10/11/2012 - 4:55pm

It's funny how FEI and S+P do this - and then what ends up happening conference by conference. Because all the OOC games are early in the season (mostly) the strength of schedule basically gets set early on - and in this case, very differently depending on the conference.

We talked before about integrating FCS games into the data to get a bit more connectivity; have you made any progress towards this?

3
by Will :: Thu, 10/11/2012 - 8:23pm

This same scenario unfolded a few years ago (I think 2009) with the FEI and the ACC, which was an average conference that "beat each other up."

Will

4
by Brian Fremeau :: Thu, 10/11/2012 - 8:35pm

FCS games don't really add connectivity unless you treat them as the "same" opponent. Even if I did consider every FCS opponent as equivalent to the No. 125 team, it would barely register with the top teams since the relevance factor of those games would be so low.

7
by Kal :: Thu, 10/11/2012 - 10:52pm

Yeah, sorry, they were two separate points. One was asking about how FEI rebalances between divisions once biases are set (or if it does at all); the other was asking about using FCS as more data.

And while FCS doesn't improve connectivity it can improve (or reduce) overall value of wins and losses.

8
by cfn_ms :: Fri, 10/12/2012 - 5:50pm

reduce the noise that comes from the rest of the games (more data points = less variance), but it introduces a fairly obvious amount of bias/variance. This is because how you interpret the AA data is drastically dependent on:

how you value blowouts;

how you calculate/factor in schedule (how you weight W/L vs margin/game stats vs schedule strength is something that can lead to massive variation from one system to another when it comes to rating big blowouts over atrocious opponents)

how you compare the AA opponents to each other or 1-A in general (if "AA" is all one bucket, then you get a bias favoring those who play the worst of the lot and hurting those who play relatively less awful AA's);

and probably other things that escape me.

Ultimately, I can see the argument for factoring it in, but I've elected not to for a while and I remain totally comfortable with that, even though it does sometimes lead to missed data (the very rare AA upset over a top 40 team, or just the somewhat rare AA near-upset over a top 40 team)

2
by Kal :: Thu, 10/11/2012 - 6:24pm

Also - how is garbage time determined per game?

5
by Brian Fremeau :: Thu, 10/11/2012 - 8:36pm

I calculate garbage time as a function of the score and the remaining possessions in the game. It's a retroactive designation.

6
by Kal :: Thu, 10/11/2012 - 10:51pm

So is it a subjective or objective value, then? That's what is confusing; it's tough for me to determine a rule that would (for instance) rule that 50-13 Oregon over Arkansas State was the end of non-garbage time (which was at halftime) vs. 52-14 was the end of non-garbage time vs. Washington (which was well into the 4th quarter and after we had substituted a number of players on both sides).